Research And Implementation Of Engine AR Assisted Maintenance Technology Driven By Knowledge Graph

As the core component of automobile,engine has the characteristics of various types,complex structure and large number of parts,its maintenance is the key and difficult point in the process of automobile maintenance.In the maintenance of engine failure,there are some problems,such as difficult to find the cause of failure,inaccurate cognition of parts structure,nonstandard maintenance operation process and low maintenance efficiency and so on.It is difficult to ensure the accurate and efficient maintenance demand of engine fault only relying on the maintenance experience of maintenance personnel and searching the maintenance manual and other traditional methods.At present,engine fault maintenance mainly relies on the computer diagnosis software fault and the maintenance personnel to find hardware fault according to the fault phenomenon and experience.However,there are few researches on the use of other technologies to assist engine fault maintenance,which can not well solve the problem of nonstandard operation and low efficiency of engine fault maintenance.This paper takes the engine as the research object,aiming at the difficulty in finding the cause of engine fault,the knowledge graph of engine fault maintenance was constructed to realize the unified structured representation of maintenance data,which provides data support for the search of fault causes and the push of maintenance resources.In view of the difficulty of cognition of parts structure and nonstandard operation in the process of engine maintenance,this paper studies the method of using AR technology to assist maintenance personnel to solve engine faults,so as to ensure the standardization and efficiency of fault maintenance.The engine fault auxiliary maintenance system is designed and developed by integrating knowledge graph and AR technology.Firstly,a large number of maintenance data generated in the process of engine fault maintenance are analyzed and processed.Through knowledge extraction,knowledge fusion,knowledge processing and other technologies,engine fault maintenance knowledge graph is constructed,and the unified structured representation of maintenance data is realized.Then,Based on the constructed knowledge graph of engine fault maintenance,the retrieval of engine fault maintenance methods was realized through the process of subgraph matching and similarity calculation.Based on the ontology model of maintenance tools in the knowledge graph,the method of matching and pushing of maintenance tools is mainly studied.Then,aiming at the key technologies needed in AR auxiliary maintenance process,the 3D tracking registration method of engine fault maintenance virtual model based on orb and KLT algorithm is studied,including feature point detection,feature description,feature point matching tracking and other processes.The 3D tracking registration of fault maintenance virtual scene and real world is realized,and the virtual reality fusion in auxiliary maintenance process is completed Design of display and human computer interaction.Finally,taking engine air filter fault maintenance as an example,the engine fault auxiliary maintenance system is built,which realizes fault retrieval,maintenance step push,maintenance action animation display and other functions,and verifies the feasibility and practicability of the proposed method.The research method in this paper provides a new idea for engine fault maintenance,has a certain practical significance to improve the accurate and efficient maintenance of engine fault,and provides a reference experience for the auxiliary maintenance process of equipment in other fields.